The purpose of this case-control study is to assess the contributions of the glutathione S-transferase (GST; genes and the cytochrome P-450 (CUP) genes to the genetic susceptibility of brain tumors and to study the effect of these genetic polymorphisms on the treatment outcome and toxicity of brain tumors. Studies to date suggest a moderate to large effect for GSTTl in selected tumor subtypes. We are targeting obtaining data on approximately 1000 cases and controls, with an adequate number of glioblastomas (n=430), astrocytomas (n=300), oligodendrogliomas (n=100) and meningiomas (n=l50) to evaluate this hypothesis with reasonable power for relative risks previously reported in the literature for GSTTl. All study subjects will be asked to complete a computer assisted self-administered questionnaire and to provide blood and/or buccal cell samples for DNA analysis. An estimated 770 recently diagnosed cases (within three months of initial diagnosis) will be recruited by research nurses from Duke University Cancer Center and 220 cases in Evanston Hospital in Evanston, IL over a four year period. An equal number of friend controls will be identified by asking each case for the name of five friends of approximately the same age (+1- 5 years), gender and race. These controls will be recruited by the research nurse and given the option of completing the protocol during an appointment at the respective clinic or using a distance based protocol. Medical records will be reviewed to obtain the clinical and treatment information for the survival and toxicity portions of the study. Linkage with the National Death Index in Year four will be completed to obtain vital statistics information and provide up to three years of follow-up time in which to assess outcomes. Study coordination and survey data management will be completed at the University of Illinois at Chicago; molecular analysis will be conducted in the laboratory of Dr. Ali-Osman at M.D. Anderson Cancer Center. The Duke Cancer Center will facilitate the neuropathology review, provide some local data collection and management support and store the remaining specimens in their central repository. Data analysis files will be compiled at UIC and statistical analysis will be completed using conditional logistic regression models as appropriate to each specific aim.